Demand for computational resources from organizations and individuals continues to grow. Increasingly, data centers meet this demand by providing large amounts of data processing and data storage capacity in a concentrated physical space. Some data centers may provide computing as a service, giving customers the flexibility to consume computational resources on an as-needed basis while sparing the customers the financial and administrative burdens of maintaining the underlying physical computing infrastructure. Other data centers may provide large-scale computation for a single organization. In either case, the optimization of data centers becomes increasingly central to computation as data centers grow and proliferate.
Optimizing data centers involves many factors that interrelate in complex ways. For example, computing devices generate waste heat, which may interfere with device performance or reliability. However, dispersal of waste heat from one device may interfere with the operation of other devices in the same data center. In addition, waste heat may make a data center environment unsuitable for human presence, thereby potentially interfering with the work of data center technicians and administrators. Furthermore, cooling systems may impose significant operating costs in the form of energy consumption, as well as costs for acquisition and maintenance.
Traditional data center layouts may place data center equipment in long rows. Unfortunately, disposing of long aisles of waste heat generated in this arrangement may pose significant ongoing expenses for energy and equipment. In some cases, traditional approaches may involve building separate walled-off areas to house some heat-producing devices, posing increased construction costs, requiring increased physical plant footprint, and/or potentially undermining other data center layout criteria (e.g., accessibility, compact infrastructure, and/or optimizing device placement for efficient communication).
Due to the many tradeoffs involved in data center designs and the highly commoditized nature of computing, an improvement to data center design can unlock a tremendous amount of value for data center operators and consumers.